Cloud computing continues to grow at a breakneck pace, yet organizations increasingly need help managing cloud costs despite its promises of being a cheaper alternative to traditional infrastructure.
The cloud changed how organizations leverage compute and storage to support and improve their businesses. The availability and rapid deployment of new resources have reduced procurement times from months to seconds, and the democratization of the provisioning process has greatly increased the growth and consumption of cloud resources. Simply put, leveraging compute and storage resources has become far more efficient.
This is akin to what happened during the Industrial Revolution. During the 19th century, the invention of the steam engine greatly increased the efficiency of consuming coal, leading to a massive uptick in coal consumption. This phenomenon is known as the Jevons Paradox, which states that increasing the efficiency of a resource will lead to increased consumption. We see this phenomenon come to life as cloud spending continues to grow at a breakneck pace, with Gartner projecting a 20.7% increase in spending in 2023.
Key reasons for inadvertent spending on cloud resources
The increased consumption of cloud resources is not without its challenges. By now, most of those who have spent time working on cloud infrastructure at large organizations have heard about or experienced an incident where the organization inadvertently spent money on resources. It has even been surveyed that 33% of cloud spend is wasted. It is easy to point to technical reasons such as over-provisioning and failing to delete stale resources; however, it is also important to truly understand why those events occurred in the first place and did not happen previously.
Shifting to the cloud has completely changed the pricing model for compute and storage. Previously business cases were created, and expenses were known prior to procuring anything. Cloud providers use a different model, Usage-Based Pricing, where resources are charged for in arrears, and that results in far slower feedback loops after the spend has been made.
This has led to many of the horror stories we hear about today, where significant expenditures have been made, and the resulting bill elicits significant sticker shock. One example of this even saw an Adobe development team waste half a million dollars in one week when they failed to detect that an extremely overprovisioned workload was costing them roughly eighty thousand dollars per day. The Usage Based Pricing model though is only part of the equation.
Lack of default traceability of expenses incurred
Another key element is examining how this model enables these situations to occur. Employees are not maliciously spending large dollar amounts on cloud resources intentionally; however, the default lack of traceability and governance when resources are created in cloud environments leads to the Tragedy of the Commons. Effectively, when there is a shared resource, people are less incentivized to be good stewards when using it.
Think of a dog park. It’s a shared resource that folks use with their pups. If you were to compare that to someone’s private backyard, which do you think would be cleaner? Part of that would undoubtedly be due to the larger volume of traffic in the dog park, but certainly, there is also an aspect of how responsible one feels for the park. If that were not the case, there would be little need for large signs indicating hefty fines for not picking up after their dogs.
The same applies to the cloud. Unless accountabilities are created around the costs that are incurred, intentionally or not, then stories like what befell Adobe will continue to happen.
“Responsibility equals accountability equals ownership. And a sense of ownership is the most powerful weapon a team or organization can have.”Pat Summitt, Legendary NCAA Basketball Coach
A sustainable solution to effectively managing costs in the cloud requires technology, people, and process
Typical attempts to solve this problem involve different approaches like procuring a SaaS tool or creating a tagging taxonomy for the organization; however, those solutions, although valuable, still need to completely address the issue as they are only focused on the technical perspective. Thoroughly addressing this problem requires addressing it from a technical perspective, and a business architecture perspective. Despite cloud computing not being new, many organizations still need to reconcile that moving to the cloud will result in new capabilities and skills that need to be gained and new processes to support the changes to how compute and storage are procured.
Addressing the people and process elements is integral to creating a lasting solution that creates organizational transparency and accountability. Failure to do so will only provide a short-term salve instead of a long-term sustainable solution to effectively managing costs in the cloud.
Three Questions Every Tech Leader Needs to Consider
As organizations grapple with this challenge, here are three questions to consider:
What are our leading cloud cost drivers today?
Organizations with multi-million-dollar cloud bills contend with hundreds of thousands of resources, incurring costs daily. How can we canvas our existing cloud footprint to see through the noise to prioritize and create an action plan for low-hanging fruit?
Do we have processes to support cost traceability and accountability as new workloads enter the cloud?
Managing cloud costs is as much about spending today as it is in the future. How can we be sure that any net-new work in the cloud doesn’t experience the same challenges we face today?
Have we equipped our employees with the right training?
With more power to support rapid innovation and development in the cloud, have we provided our employees with training and guidance on how their decisions directly impact cloud spending?